Could Elevated CO2 Lead to Over-Application of Phosphate Fertilizer to Pastures? Implications for P-Limited Ecosystem Responses to eCO2

In a recent issue of Global Change Biology, Beechey-Gradwell et al. (2025) provide compelling evidence of an elevated CO₂ (eCO₂)-driven reduction in the efficacy of phosphorus (P) fertilisation of a grassland agroecosystem. Intriguingly, the work also suggests an eCO₂ stimulation of compensatory mechanisms that allow maintenance of plant P nutrition despite the reduction of plant available P in soil. The authors argue that these findings necessitate a reassessment of whether agricultural nutrient availability tests will be fit for purpose in a high CO₂ future, and suggest a recalibration of such measurements. Furthermore, the findings of Beechey-Gradwell et al. (2025) have parallels with recent eCO₂ studies on P-limited ecosystems (e.g., Keane et al. 2023; Jian et al. 2024) that suggest eCO₂ reduces plant-available P, likely through stimulation of microbial immobilisation of P, with consequences for ecosystem productivity and its feedback to climate regulation.

As atmospheric CO₂ concentrations rise, the stimulation of greater rates of photosynthesis in plants can increase plant productivity and, consequently, increase ecosystem absorption of CO₂. Understanding this “CO₂ fertilization effect” is critical for us to determine how agricultural and natural ecosystems will respond to rising CO₂, and to quantify how much the increased sequestration of C can help slow the rise of atmospheric CO₂ and global warming.

A key constraint to the CO₂ fertilization effect on plant productivity is the availability of nutrients in soil (Terrer et al. 2019). In the case of agricultural systems that are (co-)limited by soil P availability, phosphate fertilizer may be needed to realize the CO₂ fertilization benefit. In contrast, should eCO₂ stimulate immobilization of P in soil, this may limit the effectiveness of the P fertilizer, meaning eCO₂ could create the need for more fertilizer addition along with the associated costs to farmers and increased environmental risk. The need to understand the interaction between eCO₂ and P fertilizer application is therefore clear, and this need is heightened given the globally depleting and finite resource of P for use in fertilizers.

To address this, Beechey-Gradwell et al. (2025) investigated P fertilizer and eCO₂ effects on a species-rich pasture in New Zealand. The experiment used the approach of Free-Air-CO₂-Enrichment (FACE) where a network of pipes delivers a high CO₂ environment outside on the pasture ecosystem, providing the benefit of realism. In this case, the added benefit of FACE was that it allowed the pasture to be grazed by sheep as it would in real-world conditions; this being essential for this study given their influence on P cycling through grazing and dung inputs.

Beechey-Gradwell et al. (2025) found that eCO₂ significantly reduced plant available P in soil and reduced the extent to which P fertilizer application was able to increase that P pool. The mechanism for this appeared to be rapid biological immobilisation of the P, as indicated by the accumulation of organic P in the topsoil under eCO₂ at the expense of inorganic fractions. This raises initial concerns because it suggests a negative effect of eCO₂ in that it could diminish the benefit of P fertilizer application to plant growth. However, this study found that P concentration and uptake in pasture biomass were not affected by eCO₂, suggesting eCO₂ provides an opportunity for a biological response that maintains plant P uptake or that the P fertilizer application was still enough to stop any P limitation of the pasture system. Supporting the notion that compensatory biological mechanisms may maintain plant P nutrition under eCO₂, the productivity of “P demanding” legumes (white clover) increased with eCO₂, despite the decline in available P. Therefore, as the authors point out, if biological mechanisms exist that could compensate for eCO₂-driven reductions in soil P and these are not considered in soil P tests, farmers may make unnecessary decisions to apply more P, leading to over application of the finite fertilizer resource, with its associated financial and environmental cost.

Most eCO₂ work to date has been undertaken on ecosystems where nitrogen (N), not P, is the most limiting nutrient, yet globally between a third and a half of terrestrial ecosystems are limited by P (Goll et al. 2012; Du et al. 2020). This work is therefore of further importance in adding to our growing understanding of how eCO₂ interacts with plant P availability and influences responses of P limited ecosystems. Consistent with the work of Beechey-Gradwell et al. (2025), some of the other (few) FACE studies on P-limited ecosystems are providing increasing evidence of below ground control on plant P nutrition responses—and hence productivity responses—to eCO₂. For instance, a recent FACE study on limestone and acidic grasslands found that the ability of the grasslands to increase biomass in response to eCO₂ was controlled by competition between soil microbes and plants for the limiting P resource (Keane et al. 2023). Plants effectively maintained competition for P with microbes in the limestone grassland and were able to increase productivity under eCO₂, while in the acidic grassland, microbes won out in the battle for P, leading to greater P limitation of plant growth and a surprising decline in plant productivity under eCO₂ (Keane et al. 2023). Similarly, in a FACE study on mature Eucalyptus forest, soil microbial competition for P likely restricted the P available for plant uptake and hence the capacity of the forest to sequester more C under eCO₂ (Jiang et al. 2024).

As Jiang et al. (2024) point out for P-limited forests, biological mechanisms that stimulate microbial P cycling and plant P nutrition may therefore be needed for increased C accumulation into new biomass. Beechey-Gradwell et al. (2025) suggest such mechanisms exist in their pasture system allowing maintenance of plant P nutrition under eCO₂ despite the decline in plant available P. Early evidence of such a mechanism comes from the work of Stöcklin et al. (1998) who found in a P-limited calcareous grassland that eCO₂ increased belowground biomass, with the suggestion that plants were allocating the additional C supply into roots for greater P capture. This also has parallels with the work of Taylor et al. (2024) who proposed that the increase in P-limited limestone grassland biomass in response to eCO₂ (observed by Keane et al. 2023) could be partially driven by the success of sedges (Carex) that have specialist dauciform roots. These roots could be allowing sedges (and possibly neighbouring plants) to acquire more P (Shane et al. 2006), as sedges may use the eCO₂ to increase their dauciform root function (as evidenced in pot based studies; Ballard 2001); hence driving more P uptake and ultimately greater biomass production.

Ultimately, our understanding of P-limited ecosystem responses to eCO₂, whether in agricultural or (semi-)natural ecosystem contexts, is going to need greater understanding of the interaction between eCO₂ and soil P cycling, and of how this influences plant-microbe competition for P.

Gareth K. Phoenix: conceptualization, writing – original draft, writing – review and editing. Christopher R. Taylor: conceptualization, writing – review and editing.

The authors declare no conflicts of interest.

This article is a Invited Commentary on Beechey-Gradwell et al., https://doi.org/10.1111/gcb.70150